Indicators: KAMA

[MetaQuotes]

KAMA:

Kaufman’s Adaptive Moving Average

Author: Walter

[[Deleted]]

I think you got something wrong here...
Coefficients used for calculation of filters should always sum up to 1.
When you calculate the output of this filter you use:

KAMA[i] = KAMA[i+1] + sc * (Close[i] - KAMA[i+1]);

...it should be instead:

KAMA[i] = (1-sc) * KAMA[i+1] + sc * (Close[i] - KAMA[i+1]);

Cheers,
Zyp ;)

[William Roeder]

KAMA[i] = KAMA[i+1] + sc * (Close[i] - KAMA[i+1]);

...it should be instead:

KAMA[i] = (1-sc) * KAMA[i+1] + sc * (Close[i] - KAMA[i+1]);

Wrong. The formula is

KAMA[i] = (1-sc) * KAMA[i+1] + sc * Close[i]

but that has significant problems with floating point round off. The first equation is equivalent but without the problems.

[Ivan Kornilov]

Thanks!